Add model card for CubiD

README.md

@@ -0,0 +1,27 @@
+---
+pipeline_tag: unconditional-image-generation
+---
+
+# CubiD: Cubic Discrete Diffusion for High-Dimensional Representation Tokens
+
+CubiD is the first discrete generation model designed for high-dimensional representations (e.g., 768-1024 dims). It introduces fine-grained cubic masking throughout the high-dimensional discrete representation, allowing the model to learn rich correlations across both spatial and dimensional axes.
+
+## Resources
+
+- **Paper:** [Cubic Discrete Diffusion: Discrete Visual Generation on High-Dimensional Representation Tokens](https://huggingface.co/papers/2603.19232)
+- **GitHub Repository:** [https://github.com/YuqingWang1029/CubiD](https://github.com/YuqingWang1029/CubiD)
+
+## Abstract
+
+Visual generation with discrete tokens has gained significant attention as it enables a unified token prediction paradigm shared with language models, promising seamless multimodal architectures. However, current discrete generation methods remain limited to low-dimensional latent tokens (typically 8-32 dims), sacrificing the semantic richness essential for understanding. In this paper, we present Cubic Discrete Diffusion (CubiD), the first discrete generation model for high-dimensional representations. CubiD performs fine-grained masking throughout the high-dimensional discrete representation -- any dimension at any position can be masked and predicted from partial observations. This enables the model to learn rich correlations both within and across spatial positions, with the number of generation steps fixed at $T$ regardless of feature dimensionality.
+
+## Citation
+
+```bibtex
+@article{wang2025cubic,
+  title={Cubic Discrete Diffusion: Discrete Visual Generation on High-Dimensional Representation Tokens},
+  author={Wang, Yuqing and Ma, Chuofan and Lin, Zhijie and Teng, Yao and Yu, Lijun and Wang, Shuai and Han, Jiaming and Feng, Jiashi and Jiang, Yi and Liu, Xihui},
+  journal={arXiv preprint arXiv:2603.19232},
+  year={2026}
+}
+```